Bad days for Genta and Allos, as the FDA first made their advisory committee minutes public on Friday, then turned down their drug applications yesterday. Both are nominally cancer therapeutics, and both were seen as tests of a supposedly more open attitude at the FDA, but they're very different situations.

Genta is, at first glance, a time warp of a company. They (and some others, like Isis) have been trying for years to get antisense DNA therapeutics to work. It hasn't been a very rewarding area, although most anyone who've been in the field for ten or fifteen years can recall when it was so hot you couldn't stand next to it. Delivering the DNA (or DNA analogs) has been hard, formulating them has been hard, and showing that they do anything has been very hard indeed. That was Genta's big problem this week. And that means that is was also, to a lesser extent, their partner Aventis's problem. I hope Sanofi wasn't counting on this one, but since I don't really understand why Sanofi went after Aventis at all, who can say?

The compound, Genasense, goes after the production of a protein called Bcl-2, about which more information than you want is here. A lack of Bcl-2 would make cells more vulnerable to programmed cell death (apoptosis - pronounced, if you're as much of a pain as I am, as "ay-po-tosis", not "ay-pop-tosis".) A major thing cancer cells manage to do is to bypass the you're-a-mutant-kill-yourself apoptosis signal and keep on growing, so this is a target of great interest.

But Bcl-2 works through interactions with other proteins, so it doesn't have any real small-molecule binding regions. That makes it a tough target for drug therapy (although that hasn't stopped people from trying.) Antisense or siRNA techniques could potentially stop production of the protein in the cell, though, and provide a completely new cancer therapy. Enter Genasense - and exit it, too, because it doesn't work well enough on its own to be a monotherapy, and didn't meet statistical significance as additional chemotherapy in a trial of over 700 patients.

Genta believes that a subset of the treatment group, the ones who received the compound the longest, showed enough of an effect to approve the compound, though, along with other data on things like mean tumor size. The FDA clearly didn't agree, pointing out the drug's significant toxicity, which decision should throw some cold water on the people who were hoping that things were going to be approved more easily now, with less rigorous efficacy data.

Allos got the cold shower too, but for a different sort of compound. Their RSR13 is a radiation sensitizer, working on the principle that much radiation damage in cells is done through oxidative free radicals. Unfortunately, many solid tumors are rather oxygen-deprived, due to poor circulation, so the Allos approach was to increase the rate of oxygen release from hemoglobin in general. Makes me wonder if they're going to start finding this stuff in the blood of bicycle racers.

Allos went after brain metastases in advanced breast cancer, a patient population that needs all the help it can get. Unfortunately, they didn't reach significance, either, but countered that, you guessed it, a subgroup showed a much larger effect. They might have a point, although it's always risky to ex-post-facto your clinical data. We had a discussion about this on this site a while back, and it was pointed out that for a sufficiently large set of sliced-and-diced subgroups, the real surprise would be if one of them didn't show an outlier effect.

Of the two drugs, though, I find the RSR13 story to have (at least as far as I can tell) a better chance of actually coming true after another round of clinical trials. Antisense worries me, at least in its present incarnation, and I'm not sure how many more forms it's going to get a chance to take. Both stocks were hammered on Friday and Monday, and (truth in trading time) I picked up several hundred shares of Allos today at $2.50 for the long term.